Intelligent planning, execution, and reporting of clinical trials

Abstract

Machine learning based methods for planning, execution, and reporting of clinical trials, incorporating a patient burden index are disclosed. In one aspect, there is a method for determining a patient burden index. The method includes parsing a protocol for a clinical trial. The method further includes providing factor data for each of a plurality of patients. The method further includes calculating a patient burden index for each of the plurality of patients based on the parsed protocol and the provided factor data for each of the plurality of patients.

What is claimed is: 
     
       1. A method comprising:
 establishing, by one or more processors, a model trained using historical data of patients of one or more clinical trials, wherein the historical data comprises historic patient factor data of a first set of patients and historic patient burdens of the first set of patients, the model configured to receive as input factor data for one or more patients and provide as output a patient burden index for each of the one or more patients, wherein the model is trained by:
 inputting the historic patient factor data into the model and executing the model to output predicted patient burdens for the first set of patients; and 
 adjusting one or more weights or parameters of the model based on a difference between the predicted patient burdens and the historic patient burdens; 
 
 analyzing, by one or more processors, a document of a protocol for a clinical trial to identify a schedule of actions to be taken in the clinical trial for a second set of patients; 
 identifying, by one or more processors, a set of factor data for each patient of the second set of patients to participate in the schedule of actions of the clinical trial; 
 inputting, by the one or more processors, the set of factor data for each patient of the second set of patients into the model; 
 receiving, by the one or more processors, output from the model identifying the patient burden index for each of the second set of patients, the patient burden index providing a quantitative measure of impact of the protocol on each of the second set of patients; and 
 identifying, by the one or more processors, a modification to the protocol to adjust the patient burden index for one or more patients of the second set of patients. 
 
     
     
       2. The method of  claim 1 , wherein the factor data comprises one or more of the following: patient data, trial cost data, trial time data, trial schedule data, and trial conduct data. 
     
     
       3. The method of  claim 1 , further comprising: generating, by the one or more processors, a rule that associates the factor data with the patient burden index. 
     
     
       4. The method of  claim 1 , further comprising: sending, by the one or more processors to a user, a notification based on the schedule of actions. 
     
     
       5. The method of  claim 4 , wherein sending the notification is responsive to determining that an action of the schedule of actions is not completed at a time according to the schedule of actions. 
     
     
       6. The method of  claim 4 , wherein the notification comprises:
 a notification to send, to one or more patients of the second set of patients, materials for the clinical trial; 
 a notification to adjust the schedule of actions based on an event; 
 a notification of an event; or 
 a notification to remind one or more patients of the second set of patients to complete an action of the schedule of actions. 
 
     
     
       7. The method of  claim 1 , further comprising:
 storing, by the one or more processors in a database, an indication of an event responsive to determining that an action of the schedule of actions is not completed at a time according to the schedule of actions; and 
 determining, by the one or more processors, a recommendation for the event. 
 
     
     
       8. The method of  claim 7 , wherein the event comprises:
 an adverse event; 
 a treatment-emergent event; or 
 a protocol deviation. 
 
     
     
       9. The method of  claim 1 , further comprising:
 updating, by the one or more processors, sets of factor data for second one or more patients of the second set of patients based on one or more events; and 
 executing, by the one or more processors using the updated sets of factor data as input, the model to calculate a second patient burden index for each patient of the second one or more patients of the second set of patients. 
 
     
     
       10. The method of  claim 9 , wherein the one or more events comprises at least one of a reminder, an action of the schedule of actions, or failure to perform an action of the schedule of actions. 
     
     
       11. The method of  claim 1 , wherein identifying the schedule of actions further comprises: calculating the schedule of actions based on an algorithm, the algorithm comprising either a greedy algorithm or an evolutionary algorithm. 
     
     
       12. The method of  claim 1 , further comprising:
 analyzing, by the one or more processors during the clinical trial, one or more events; and 
 determining a likelihood of meeting one or more endpoints of the clinical trial by the end of the clinical trial. 
 
     
     
       13. A system, comprising:
 one or more processors, coupled to memory, to: 
 establish a model trained using historical data of patients of one or more clinical trials, wherein the historical data comprises historic patient factor data of a first set of patients and historic patient burdens of the first set of patients, the model configured to receive as input factor data for one or more patients and provide as output a patient burden index for each of the one or more patients, wherein the model is trained by:
 inputting the historic patient factor data into the model and executing the model to output predicted patient burdens for the first set of patients; and 
 adjusting one or more weights or parameters of the model based on a difference between the predicted patient burdens and the historic patient burdens; 
 
 analyze a document of a protocol for a clinical trial to identify a schedule of actions to be taken in a clinical trial for a second set of patients;
 identify a set of factor data for each patient of the second set of patients to participate in the schedule of actions of the clinical trial; 
 input the set of factor data for each patient of the second set of patients into the model; 
 receive output from the model identifying the patient burden index for each of the second set of patients, the patient burden index providing a quantitative measure of impact of the protocol on each of the second set of patients; and 
 identify a modification to the protocol to adjust the patient burden index for one or more patients of the second set of patients. 
 
 
     
     
       14. The system of  claim 13 , the one or more processors further to: generate a rule that associates the factor data with the patient burden index. 
     
     
       15. The system of  claim 13 , the one or more processors further to: send a notification based on the schedule of actions and responsive to determining that an action of the schedule of actions is not completed at a time according to the schedule of actions. 
     
     
       16. The system of  claim 13 , further comprising:
 storing, by the one or more processors in a database, an indication of an event responsive to determining that an action of the schedule of actions is not completed at a time according to the schedule of actions; and 
 determining, by the one or more processors, a recommendation for the event. 
 
     
     
       17. The system of  claim 13 , further comprising:
 updating, by the one or more processors, sets of factor data for second one or more patients of the second set of patients based on one or more events; and 
 executing, by the one or more processors using the updated sets of factor data for the second one or more patients of the second set of patients as input, the model to calculate a second patient burden index for each patient of the second one or more patients of the second set of patients. 
 
     
     
       18. The system of  claim 13 , further comprising:
 analyzing, by the one or more processors during the clinical trial, one or more events; and 
 determining a likelihood of meeting one or more endpoints of the clinical trial by the end of the clinical trial. 
 
     
     
       19. A non-transitory computer readable storage medium comprising instructions stored thereon that, when executed by a processor, cause the processor to:
 establish a model trained using historical data of patients of one or more clinical trials, wherein the historical data comprises historic patient factor data of a first set of patients and historic patient burdens of the first set of patients, the model configured to receive as input factor data for one or more patients and provide as output a patient burden index for each of the one or more patients, wherein the model is trained by:
 inputting the historic patient factor data into the model and executing the model to output predicted patient burdens for the first set of patients; and 
 adjusting one or more weights or parameters of the model based on a difference between the predicted patient burdens and the historic patient burdens; 
 
 analyze a document of a protocol for a clinical trial to identify a schedule of actions to be taken in a clinical trial for a second set of patients; 
 identify a set of factor data for each of the second set of patients to participate in the schedule of actions of the clinical trial; 
 input the set of factor data into the model; 
 receive output from the model identifying the patient burden index for each of the second set of patients, the patient burden index providing a quantitative measure of impact of the protocol on each of the second set of patients; and 
 identify a modification to the protocol to adjust the patient burden index for one or more patients of the second set of patients. 
 
     
     
       20. The medium of  claim 19 , wherein the instructions stored thereon further cause the processor to: generate a rule that associates the factor data with the patient burden index.